Alibaba Releases MiniMax-H3 Text-to-Video Model with ControlNet

alibaba-pai · hf · 2026-08-25

Alibaba released the MiniMax-H3-Fun-Controlnet-Union model on Hugging Face, a text-to-video pipeline tool. The model supports various generation modes including video-to-video and image-text-to-video, integrating ControlNet technology for enhanced control.

Related event: Alibaba PAI Open-Sources Unified ControlNet for MiniMax-H3 Video Generation(5 posts)→

Original post →

More from Multimodal

Multimodal channel →